Doug Coleman
45ed9b3b80
use -fPIC on all platforms for ffi lib
2009-03-27 15:39:13 -05:00
Doug Coleman
785e3bcfd0
only use -fPIC on the ffi library
2009-03-26 10:47:25 -05:00
Doug Coleman
77fccb2782
build with -fPIC on linux64
2009-03-26 10:38:00 -05:00
Doug Coleman
97283b9857
compile an ffi test library, reduce factor runtime by 6kb!
2009-03-25 13:06:20 -05:00
sheeple
c6837fbe3e
Update cpu-ppc.S for quotation layout change
2009-03-24 01:07:38 -05:00
Slava Pestov
2a7848053f
Fix 64-bit build issue; relocation entries are 32-bit on all platforms
2009-03-19 23:21:32 -05:00
Slava Pestov
ae09d85d84
More compact relocation layout
2009-03-19 20:03:07 -05:00
Slava Pestov
8f059e07a7
Move flags from F_CODE_BLOCK to F_BLOCK for further space savings
2009-03-19 03:49:49 -05:00
Slava Pestov
78f168e304
Remove code_length field from F_CODE_BLOCK struct, and make F_BLOCK the head of F_CODE_BLOCK to simplify other code
2009-03-19 03:45:37 -05:00
Slava Pestov
44d61b71f6
Store next_free pointer to free blocks only
2009-03-19 03:00:21 -05:00
Slava Pestov
be4fb1e7d9
Move call( and execute( to core
2009-03-16 20:11:36 -05:00
Slava Pestov
1559b74640
Add new check-datastack primitive and re-implement call( with it, instead of using with-datastack. call( is now 5x faster
2009-03-16 06:16:51 -05:00
Slava Pestov
a2d93dad1d
Fix to make new_ui compile with X11 libraries on Mac OS X
2009-03-05 13:38:55 -06:00
Slava Pestov
2fee574b4f
Working on X11 backend
2009-03-02 02:55:54 -06:00
Slava Pestov
4257cd55e0
fix problem with bignum-bit? -- return value would be truncated if sizeof(int) != sizeof(bignum_digit_type)
2009-02-22 19:45:05 -06:00
Joe Groff
5dd4bbcf42
Defuse RTLD_GLOBAL time bomb in os-unix.c
2009-02-10 19:58:53 -06:00
Doug Coleman
201296c040
dllexport err_no and clear_err_no
2009-02-06 18:46:23 -06:00
sheeple
173b0ee78d
Add some more tests for complex numbers in FFI
2009-02-06 12:21:53 -06:00
Slava Pestov
05632b8525
Don't use complex.h since *BSDs don't have it in latest release versions (gah!); add DLLEXPORT for ffi_test_45 to make it work on Windows
2009-02-06 11:03:52 -06:00
Slava Pestov
f9bc9a3198
Fix VM compile error
2009-02-06 04:53:08 -06:00
Slava Pestov
7bb0e78314
Add support for C99 complex float and complex double types to FFI
...
They are named complex-float and complex-double in the Factor world
2009-02-06 04:02:00 -06:00
unknown
b6c3bc892d
Fix the build support for windows again, have the makefile load Config.* files in the top level so that freetype.dll and zlib1.dll targets can make
2009-02-04 18:21:58 -06:00
Doug Coleman
eec86d6043
Fix io.launcher on Windows -- Microsoft does the same thing with devenv
2009-02-04 13:05:52 -06:00
Doug Coleman
a60e11b89e
add RTLD_GLOBAL to dlopen flags. load atlas before cblas on freebsd
2009-01-30 16:45:30 -06:00
Doug Coleman
ed7b49df78
Windows vm now tries vm-console.image or vm.image first, then removes -console if it exists and tries to find the image again
2009-01-29 19:57:01 -06:00
Doug Coleman
8acba1543a
oops, didnt mean to remove these two lines
2009-01-29 13:38:07 -06:00
Doug Coleman
2d960f4103
Create two binaries: factor.com and factor.exe. factor.com is the console version, factor.exe is win32. also make a factor shell script called factor to help cygwin out because it ignores the env vars
2009-01-29 13:15:16 -06:00
Slava Pestov
c074740746
Split off data_heap.c from data_gc.c; split off write_barrier.h, local_roots.h from data_gc.h, misc other cleanups
2009-01-24 23:39:00 -06:00
Slava Pestov
ed1a2855b5
Big code GC refactoring
...
- Move code block-specific parts of code_gc.c to code_heap.c
- code_gc.c is now a generic malloc-style heap
- New code_block.c to split up large code_heap.c
- Clean up relocation code and remove some duplication
- Fix problems with code heap not being updated properly with uninterned words (bug re
ported by doublec)
- Remove boolean parameter from modify-code-heap primitive
- Less unnecessary fixup speeds up bootstrap slightly
2009-01-24 20:13:17 -06:00
Slava Pestov
670d0106d1
Clean up some duplication
2009-01-24 17:42:28 -06:00
Slava Pestov
32bb531621
Move literal tables out of code heap, instead each code block holds a reference to the literal table
2009-01-24 17:01:01 -06:00
Slava Pestov
a8d1459c08
Rename compiled slot of words to optimized, to reflect what it really means (all words are compiled)
2009-01-23 00:37:02 -06:00
Doug Coleman
09b82321b5
remove failed proper-save from vm
2009-01-21 23:32:23 -06:00
Doug Coleman
1dc7ecfdc5
lol the hak...fix coming up
2009-01-21 20:12:09 -06:00
Doug Coleman
a9277c71fd
fix macosx initialization of executable_path
2009-01-21 18:07:26 -06:00
Doug Coleman
99cd664961
better save_image that doesn't corrupt your factor image if you're out of disk space. should throw exception instead of printing error messages if saving fails..
2009-01-21 15:02:27 -06:00
Doug Coleman
9f95c9611b
fix vm path on openbsd, fix compiler warning
2009-01-19 21:19:32 -06:00
Slava Pestov
0d56665522
Fix Windows build
2009-01-16 14:21:44 -06:00
Slava Pestov
07ec51b24a
Clean up embedding API
2009-01-14 18:40:23 -06:00
Doug Coleman
585d977ed0
remove a couple comments from Config.windows files
2009-01-13 20:54:39 -06:00
Slava Pestov
b2e45f199e
Calling dispose on unix stdin no longer kills the VM
2008-12-20 18:02:08 -06:00
Slava Pestov
00212bb626
Save and restore r31
2008-12-16 02:06:53 -06:00
unknown
b5d98e6a1e
Fix warning
2008-12-15 21:19:04 -06:00
U-SLAVA-DFB8FF805\Slava
f4c3bf1dd6
Fix compile warning and fix (sleep) primitive on Windows
2008-12-13 04:49:16 -06:00
Joe Groff
88856fac21
OSX VM requires function from Cocoa framework, so link against it even when building for X11
2008-12-10 12:22:17 -08:00
Slava Pestov
ec23584f48
Finish up O(1) byte array allocation
2008-12-09 18:17:04 -06:00
Slava Pestov
eb79c6ab71
Fix conflict
2008-12-09 17:52:45 -06:00
Slava Pestov
9ad51490af
Fix GC crash with -generations=2
2008-12-09 16:26:30 -06:00
Slava Pestov
833d9f9c0b
Fix quotation pooling
2008-12-08 21:24:45 -06:00
Slava Pestov
e4f8448eb1
Fix some problems with arithmetic type inference, exposed by recent changes to log2 word
...
- declared input type for bignum-shift was stricter than the runtime behavior, leading to bad propagation of type info if shift count was a bignum
- types inferred for type functions which used number-valued/integer-valued/real-valued were not always precise, eg bignum bignum bitxor => integer
- add interval-log2, type function for (log2)
- remove math-class-min, it was useless
2008-12-07 19:44:49 -06:00